Starting Monday, October 5, 2020, the NVIDIA GPU Technology Conference (GTC) is offering online sessions for you to learn AWS best practices to accomplish your machine learning (ML), virtual workstations, high performance computing (HPC), and internet of things (IoT) goals faster and more easily.
Amazon Elastic Compute Cloud (Amazon EC2) instances powered by NVIDIA GPUs deliver the scalable performance needed for fast ML training, cost-effective ML inference, flexible remote virtual workstations, and powerful HPC computations. At the edge, you can use AWS IoT Greengrass and SageMaker Neo to extend a wide range of AWS Cloud services and ML inference to NVIDIA-based edge devices so the devices can act locally on the data they generate.
AWS is a Global Diamond Sponsor of the conference.
Available sessions
The following sessions are available from AWS:
- As a deep learning developer or data scientist, you can choose from multiple GPU EC2 instance types based on your training and deployment requirements. You can access instances with different GPU memory sizes, NVIDIA GPU architectures, capabilities (precisions, Tensor Cores, NVLink), GPUs per instance, number of vCPUs, system memory, and network bandwidth. We’ll share some guidance on how you can choose the right GPU instance on AWS for your deep learning projects. You’ll get all the information you need to make an informed choice for GPU instance for your training and inference workload.
- Speaker: Shashank Prasanna, Senior Developer Advocate, AI/ML, Amazon Web Services
- Virtual workstations on AWS enable studios, departments, and freelancers to take on bigger projects, work from anywhere, and pay only for what they need. Running on Amazon EC2 G4 instances, virtual workstations employ the power of NVIDIA T4 Tensor Core GPUs and Quadro technology, the visual computing platform trusted by creative and technical professionals. Virtual workstations have become essential to creative professionals seeking cloud solutions that enable remote teams to work more efficiently, and keep creative productions moving forward. Join this session to learn more about how virtual workstations on AWS work, who is using them today, and how to get started.
- Speaker: Haley Kannall, CG Supervisor, Amazon Web Services
- We’ll discuss how we can optimize edge video inferencing performance by leveraging AWS infrastructure and NVIDA Deepstream. We’ll emphasize three major features at the edge: (1) massively deploying trained models to NVIDIA Jetson devices using AWS IoT Greengrass, (2) local communication and control between AWS IoT Greengrass engines and Deepstream applications through MQTT messaging, and (3) uploading inferencing results to the cloud for further analytics.
- Speaker: Yuxin Yang, IoT Architect, Amazon Web Services
- In this presentation, we provide an overview of AWS Wavelength, how it integrates with the Mobile Edge carrier network and improves the performance of Mobile Edge applications. Wavelength Zones are AWS infrastructure deployments that embed AWS compute and storage services within telecommunications providers’ datacenters at the edge of the 5G network, so application traffic can reach application servers running in Wavelength Zones without leaving the mobile providers’ network. Customers with edge data processing needs such as image and video recognition, inference, data aggregation, and responsive analytics can use Wavelength to perform low-latency operations and processing right where their data is generated, reducing the need to move large amounts of data to be processed in centralized locations. We deep dive into these Mobile Edge applications running at the AWS Wavelength Zones using Amazon EC2 G4 instances powered by NVIDIA T4 Tensor Core GPUs.
- Speaker: Sebastian Dreisch, Head of Wavelength GTM, Amazon Web Services
- Development of autonomous driving systems presents a massive computational challenge, including processing petabytes of sensor data, which impacts time to market, scale, and cost, throughout the development cycle. Training, testing, validating, and deploying self-driving systems requires large-scale compute and storage infrastructure to support the end-to-end workflow. AWS offers a highly scalable and reliable solution for AV development including the latest generation GPUs from NVIDIA. By attending this webinar, you will learn about AWS AV solution architectures for data ingest, data management, simulation, and distributed model training, as well as strategies for cost optimization. NVIDIA will share new details about the next generation NVIDIA Ampere (A100) architecture. Attendees will walk away with an understanding of how AWS and NVIDIA can help streamline AV development and reduce IT costs and time-to-market.
- Speakers: Shyam Kumar, Principal HPC Business Development Manager, Amazon Web Services, and Norm Marks, Global Senior Director, Automotive Industry, NVIDIA
- We’re all used to change. In business, change is often predictable—different seasons, large-scale events, and new releases all drive fluctuations we’re used to. But right now, there’s nothing normal about the changes you’re facing. The only constant is uncertainty. And uncertainty is expensive. In the absence of an omniscient crystal ball, the next best thing is cloud and ML. This presentation is going to cover how to deal with the unexpected. Whether it’s rapidly changing traffic, shifting data sources, or model drift, we’ll cover how you can better manage spikes and dips of all sizes and improve predictions with AI to maximize your efficiencies today.
- Speaker: Allie Miller, US Head of ML Business Development for Startups and Venture Capital at AWS, Amazon Web Services
Accelerating Data Science with NVIDIA RAPIDS (Scheduled session ID: A22042)
- Data science workflows have become increasingly computationally intensive in recent years, and GPUs have stepped up to address this challenge. With the RAPIDS suite of open-source software libraries and APIs, data scientists can run end-to-end data science and analytics pipelines entirely on GPUs, allowing organizations to deliver results faster than ever. The AWS Cloud lets you access a large number of powerful NVIDIA GPUs with Amazon EC2 P3 based on V100 GPUs, Amazon EC2 G4 based on T4 GPUs, and upcoming A100-based GPU instances. We’ll go through the end-to-end process of running on RAPIDS on AWS. We’ll start by running RAPIDS libraries on a single GPU instance. Next, we’ll see how you can run large-scale hyperparameter search experiments with RAPIDS and Amazon SageMaker. Finally, we’ll run RAPIDS distributed ML using Dask clusters on Amazon EKS and Amazon ECS.
- Speaker: Shashank Prasanna, Senior Developer Advocate, AI/ML, Amazon Web Services
Interactive Scientific Visualization on AWS with NVIDIA IndeX SDK (On-Demand session ID: A21610)
- Scientific visualization is critical to understanding complex phenomena modeled using HPC simulations. However, it has been challenging to do this effectively due to the inability to visualize large data volumes (> 1 PB) and lack of collaborative workflow solutions. NVIDIA IndeX on AWS, a 3D volumetric interactive visualization toolkit, addresses these problems by providing a scalable scientific visualization solution. NVIDIA IndeX allows you to make real-time modifications and navigate to the most pertinent parts of the data to gather better insights faster. IndeX leverages GPU clusters for scalable, real-time visualization and computing of multi-valued volumetric data together with embedded geometry data. We’ll demonstrate 3D volume rendering at scale on AWS using IndeX.
- Speakers: Karthik Raman, Senior Solutions Architect, HPC, Amazon Web Services, and Dragos Tatulea, Software Engineer, NVIDIA
Conclusion
You can also visit AWS and NVIDIA to learn more or apply for a free trial to use NVIDIA GPU-based Amazon EC2 P3 instances powered by NVIDIA V100 Tensor Core GPUs and Amazon EC2 G4 instances powered by NVIDIA T4 Tensor Core GPUs. Learn more about GTC on the GTC 2020 website. We look forward to seeing you there!
About the Author
Geoff Murase is a Senior Product Marketing Manager for AWS EC2 accelerated computing instances, helping customers meet their compute needs by providing access to hardware-based compute accelerators such as Graphics Processing Units (GPUs) or Field Programmable Gate Arrays (FPGAs). In his spare time, he enjoys playing basketball and biking with his family.